Changing weather patterns are becoming more common, and predicting them is becoming even more difficult. This is especially true for electric and gas utilities, which use historical weather data when developing predictions of future weather patterns, which are in turn used to project future demand and ultimately rates. In the past, public utility commissions have accepted the practice of using 30-year historical averages to develop weather normalization curves and forecasts submitted as part of the rate case process.

However, as of late, there has been discussion in the utility industry regarding the usefulness of this 30-year historical data when developing weather and demand projections. Temperature normals are becoming less accurate, with U.S. and global climates experiencing a warming trend over the past 40 years. To address this concern, there has been some movement toward having utilities use shorter-year historical averages, typically 10 or 15 years, to better reflect this warming trend in rate case filings. Though additional data is needed, utilities that use 30-year data may be under-representing future temperatures and, by default, the potential additional costs that are associated with warmer temperatures and greater variability.

Key Details

• According to a 2013 paper published in the Journal of Applied Meteorology and Climate, the use of 30-year surface temperature averages as estimates of future temperatures will, in many instances, result in a “cold bias”—predicting temperatures will be colder than those actually experienced; using the most recent 15-year average is the best method for developing weather normalization curves
• According to industry experts, more and more evidence is being submitted by utilities that shows that 30-year averages of weather data are no longer representative of future weather trends
• Recently, the New York Public Service Commission authorized the use of 10-year historical averages for the development of weather normalization calculations for rate cases submitted by Central Hudson Gas and Electric, New York State Electric & Gas, and Consolidated Edison

Implications

• Utility commissions may increase their scrutiny of historical weather data submitted as part of a rate case filing
• A utility that utilizes 30-year historical data may want to compare the accuracy of its past forecasts to recent actuals. If there is significant variation, the utility may want to determine if the variation would have occurred had it limited its trend data to more recent periods; it may be appropriate to modify the inputs to its weather models
• If a utility determines it should update its weather models (which are already recognized as an inexact science) to reflect the recent warming trends, it should be prepared to address the potential for climate-related controversy when filing its next rate case

More Information

SNL: Changing Weather Disrupting Norms in Utility Ratemaking

The Philadelphia Inquirer: Staying Normal

Journal of Applied Meteorology and Climatology: Performance of Alternative ‘Normals’ for Tracking Climate Changes, Using Homogenized and Non-Homogenized Seasonal U.S. Surface Temperatures

Itron Forecasting White Paper: Defining Normal Weather for Energy and Peak Normalization

This report is part of the Regulatory Minute series. To view all featured Regulatory Minutes, please click here.

Additional Contributing Authors: Matt Pierce, Eric Hanson